Abstract: A computer-implemented method for detecting a fraudulent payment transaction, including steps for receiving user location information, receiving user payment history information and building a predictive model, for a user, based on the user location information and the user payment history information. In certain aspects, the method further includes steps for receiving user status information in addition to a transaction event and evaluating the transaction event using the user status information and the predictive model to determine a likelihood that the transaction event is fraudulent. Systems and computer-readable media are also provided.

Abstract: The subject matter of this specification can be implemented in, among other things, a method for providing content suggestions based on an event. The method includes a step for receiving event metadata. The method also includes a step for retrieving interest profiles for a plurality of attendees or invitees from the list of attendees or invitees. The method also includes a step for determining attributes for the event based on the received event metadata. The method also includes a step for determining content suggestions based on the retrieved interest profiles and the determined attributes. The method also includes a step for providing one or more of the determined content suggestions for a user to manage the event.

Abstract: Various aspects of the subject technology relate to systems, methods, and machine-readable media for initiating a client-side user model. A system is configured to receive instructions to initiate a client-side user model for a user, wherein the instructions include a device identifier and a user identifier, identify, based on the user identifier, a user model for the user, wherein the user model is stored on a data server and comprises information about the user, transmit the information about the user in the user model on the data server to a client device corresponding to the device identifier, and delete the user model on the data server.

Abstract: Various aspects of the subject technology relate to systems, methods, and machine-readable media for providing personalized services using a client-side user model. A client device is to transmit a request for a service to a server. The server is to receive a request for a personalized service for a user, transmit, to a data server, a request for user data for the user to be used to personalize the service receive, from the data server, a notification that the user data is stored on the client device in response to transmitting the request for the user data, and provide personalization instructions and the service to the client device. The client device is further to receive the requested service and the personalization instructions from the server, personalize the service using the personalization instructions and the user model stored on the client device, and provide the personalized service to a user.